ARTICLE IN PRESS
Journal of Plant Physiology 162 (2005) 151—160
www.elsevier.de/jplph
Apple flavonols during fruit adaptation to solar
radiation: spectral features and technique for
non-destructive assessment
Mark N. Merzlyaka,, Alexei E. Solovchenkoa, Alexei I. Smaginb,
Anatoly A. Gitelsonc
a
Faculty of Biology, Department of Physiology of Microorganisms, Moscow State University,
119992 Moscow GSP-2, Russia
b
I.V. Michurin All-Russia Research Institute for Horticulture, Michurinsk, Tambov Region 393760, Russia
c
Center for Land Management and Information Technologies, School of Natural Resources,
University of Nebraska-Lincoln, 113 Nebraska Hall. Lincoln, NE 68588-0517, USA
Received 14 November 2003; accepted 7 July 2004
KEYWORDS
Flavonols;
Non-destructive
assessment;
Apples;
Reflectance;
Adaptation to solar
radiation
Summary
Spectral properties of flavonols of three varieties (Golden Delicious, Antonovka, and
Renet Simirenko) of anthocyanin-free apple fruit were investigated with reflectance
spectroscopy. The results of spectral and biochemical analyses suggested that fruit
reflectance in a broad spectral range 365–430 nm is strongly dependent on and, in
sunlit fruit surfaces, governed by flavonols. The build up of peel flavonols (mainly rutin
and other quercetin glycosides) resulted in a dramatic decrease of fruit reflectance in
this range, flattening of the spectrum, and extending the region with low reflectance
(4–5%) to ca. 410 nm. The spectral features observed suggest that flavonols contribute
significantly to screening of excessive radiation, not only UV-A, but in the short-wave
bands of chlorophyll and carotenoid absorption in the visible part of the spectrum as
well. To retrieve quantitatively flavonol content from reflectance spectra, we tested
the applicability of an inversion technique developed for non-destructive leaf pigment
assessment. The model for flavonol content assessment was suggested in the form
1
ðR1
410 R460 ÞR800 ; where Rl is reflectance at wavelength l. The model was linearly
related to flavonol content between 8 and 220 nmol/cm2 with the coefficient of
determination r2=0.92 and root mean square error of flavonol estimation of 20 nmol/
cm2 regardless of cultivar, chlorophyll, and carotenoid content.
& 2004 Elsevier GmbH. All rights reserved.
Abbreviations: Chl, Chlorophyll(s); Car, Carotenoid(s); Flv, Flavonol(s); FRI, Flavonol Reflectance Index; IP, Inflection Point; NIR, Near
Infra Red
Corresponding author. Tel: +7-095-9392587; fax: +7-095-9393807.
E-mail addresses: [email protected], [email protected] (M.N. Merzlyak).
0176-1617/$ - see front matter & 2004 Elsevier GmbH. All rights reserved.
doi:10.1016/j.jplph.2004.07.002
ARTICLE IN PRESS
152
Introduction
Flavonols (Flv) are an abundant group of phenolic
compounds involved in a number of physiological
functions in higher plants (Markham, 1989; Harborne and Williams, 2000). Several lines of evidence suggest their protective role against
damages induced by excessive UV and visible
radiation (Tevini et al., 1991; Day et al., 1993;
Bornman et al., 1997; Jansen et al., 1998; Burchard
et al., 2000; Cerovic et al., 2000; Mazza et al.,
2000; Havaux and Kloppstech, 2001; Kolb et al.,
2001; Liakoura et al., 2003). Although Flv are able
to exert their photoprotective action via several
mechanisms (Takahama, 1983; Bornman et al.,
1997; Havaux and Kloppstech, 2001), it is generally
accepted that the most important mechanism is
related to the screening of excessive solar radiation
by Flv accumulating in the surface plant tissue
structures: cuticle (Krauss et al., 1997; Solovchenko and Merzlyak, 2003) and epidermis (Tevini et al.,
1991; Day et al., 1993; Mazza et al., 2000).
Accordingly, the knowledge of Flv in vivo optical
properties as well as the possibility of nondestructive estimation of their content could give
a clue about potential of a plant organism for
acclimation to and cope with excessive solar
radiation as well as to a variety of other stresses,
which are often accompanied by the induction of
Flv biosynthesis (Harborne and Williams, 2000;
Mazza et al., 2000). The main body of evidence
for the photoprotective role of Flv and other
phenolics in plants was obtained in the experiments
with leaves (Burchard et al., 2000; Mazza et al.,
2000). However, the contribution of phenolics into
optical properties of leaves is difficult to evaluate
quantitatively due to strong interference by Chl
and Car (Cerovic et al., 2000).
Compared to leaves, apple fruits contain much
lower quantities of the pigments and possess more
resolved reflectance spectra with distinct features
attributable to Flv presented mainly by rutin and
other quercetin glycosides (Escarpa and González,
1998; Reay and Lancaster, 2001; Solovchenko and
Merzlyak, 2003). In apple fruits, strong solar
radiation induces a remarkable increase in Flv
evidently to prevent the development of a phototooxidative damage—sunscald (see Merzlyak et al.,
2002). The considerable difference in Flv content
between sunlit and shaded fruit surfaces have been
recorded through different stages of on-tree fruit
ripening and were retained in the course of
prolonged fruit storage (Merzlyak et al., 2002).
The build up of Flv in surfaces of anthocyanin-free
apples exposed to direct sun-light resulted in a
steep decrease of whole fruit reflectance in the
M.N. Merzlyak et al.
UV-A range (Solovchenko and Merzlyak, 2003). The
corresponding sun-light-induced increase in absorption in this band has been also observed in isolated
cuticles and, especially, in the peel of apple fruit
(Solovchenko and Merzlyak, 2003). The accumulation of Flv in apple peel resulted in increased
resistance of the fruit photosynthetic apparatus to
high fluxes of visible (Merzlyak et al., 2002) and UV
radiation (Solovchenko and Schmitz-Eiberger,
2003). These circumstances make apple fruits a
useful natural system for studies of the influence of
solar radiation on the expression of photoprotective reactions and, in particular, the effect of Flv on
the optical properties of plants.
Recently, a conceptual model relating remotely
sensed reflectance and pigment content was
developed and applied for non-destructive estimation of Chl, Car, and anthocyanins in higher plant
leaves (Gitelson and Merzlyak, 1996; Gitelson et
al., 1996, 2001, 2002, 2003, Merzlyak et al., 2003a)
and fruits (Merzlyak et al., 2003a, b). The model
was devised to isolate the absorption coefficient of
a pigment of interest from reflectance spectra. The
following relationship between pigment content
and reflectance was used:
1
½Pigment / ðR1
l1 Rl2 ÞRl3 :
Here l1 is the spectral region such that R1
l1 is
maximally sensitive to the absorption of the pigment
of interest, although it is still affected by the
absorption of other pigments and leaf scattering. l2
is the spectral region such that R1
l2 is minimally
sensitive to the pigment of interest and maximally
sensitive to absorption by other interfering pigment(s). It was also assumed that the absorption
coefficient of other pigments at l2 was close to that
1
at l1. Thus, the difference ðR1
l1 Rl2 Þ was related
to the content of a pigment of interest. However, it
was still affected by the variability in leaf structure
and thickness (Gitelson et al., 2003), that is, the
scattering by the medium. l3 is a spectral region
minimally affected by the absorption of pigments,
and it therefore was used to compensate for the
variability in scattering between samples.
In this study we attempted to investigate in vivo
spectral properties of Flv in apple fruit in more
detail and quantitatively. In addition, we investigated the applicability of the conceptual model for
the non-destructive estimation of Flv content. We
hypothesized that tuning of the spectral regions l1,
l2, and l3 according to the optical characteristics of
the apple fruit is the only requirement to achieve
the desired result. To the best of our knowledge,
non-destructive assessment of Flv in plants has not
been considered to date in the literature.
ARTICLE IN PRESS
Apple flavonols during fruit adaptation to solar radiation
153
Material and methods
Results
Plant material
Reflectance spectral features of flavonols
Ripe anthocyanin-free fruits of apple (Malus domestica Borkh.) varieties (Golden Delicious, Antonovka and Renet Simirenko) without symptoms of
sunscald were used in this study. Antonovka apples
were grown in commercial orchards in Michurinsk
(Tambov region, Russia); Golden Delicious and
Renet Simirenko fruits were delivered from the
Krasnodar region of Russia where accidents of
sunscald are common. All samples have been
studied within 2–3 weeks.
Fruits of all apple cultivars examined in this study
possessed a considerable variation in peel Flv, Chl
and Car (Fig. 1, Table 1). The peels of yellowish
fruits of Golden Delicious apples were characterized by a considerable level of Car and relatively
low amounts of Chl. Green in colour, Renet
Simirenko fruits possessed high Chl and Car
contents among studied varieties. Pale-green to
yellowish Antonovka apples represented an intermediate case with moderate content of both Chl
and Car (Table 1). As a rule, more green peel
(characteristic of shaded surfaces of the fruits),
along with higher Chl content, possessed lower Flv
levels, whereas higher Flv and lower Chl contents
were found in the peels from sunlit surfaces (see
also Merzlyak et al., 2002). Flv content in the peel
possessed non-uniform distribution: in the majority
of fruits from shaded and sunlit surfaces it was in
the ranges of 10–75 and 75–250 nmol/cm2, respectively (Fig. 1). Fruits of all apple cultivars studied
Spectral measurements
Whole fruit reflectance spectra were recorded with
a 150–20 Hitachi spectrophotometer equipped with
an integrating sphere against barium sulphate as a
standard. The spectra were recorded at 2-nm
sampling intervals in 400–800 nm range and in some
experiments with Antonovka fruits, in the range
300–800 nm. The acquired data were interfaced to
IBM-compatible personal computer.
Pigment analysis
The procedure allowing simultaneous quantification of Chl, Car, and Flv in an extract from apple
peel zone used for reflectance measurements was
employed essentially as described in Solovchenko
et al. (2001). Peel disks (total area of 3.8 cm2) were
homogenized in chloroform–methanol (2:1, vol/
vol) in the presence of MgO. After completion of
extraction, homogenates were filtered through a
paper filter, and distilled water (1/5 of total
extract volume) was added. Then extracts were
centrifuged at 3000g for 10 min until phase separation. Chl and Car concentrations were quantified
spectrophotometrically in lower (chloroform) phase
using coefficients reported by Wellburn (1994). The
upper (water–methanol) phase was used for assay
of Flv, which were quantified spectrophotometrically
using
molar
absorption
coefficient
e358=25.4 mM1 cm1 determined for rutin in 80%
aqueous methanol. After determination of Flv the
water–methanol phase was acidified with HCl (final
concentration of HCl=0.1%) and used for quantification of anthocyanins by measuring absorbance at
530 nm; absorption coefficient of 30 mM1 cm1
(Strack and Wray, 1989) was accepted. Flv (in
equivalent amounts of rutin) as well as other
pigment content were expressed relative to fruit
surface area.
Figure 1. Distribution of flavonol content in peel samples
from shaded (1) and sunlit (2) surfaces of fruits of apple
cultivars studied.
ARTICLE IN PRESS
154
Table 1.
M.N. Merzlyak et al.
Peel pigment content (nmol/cm2) of apple fruits
Golden delicious (n=13)
Min–Max
Average7STDa
Coefficient of variation (%)
Antonovka (n=35)
Min—Max
Average7STD
Coefficient of variation (%)
Renet Simirenko (n=29)
Min–Max
Average7STD
Coefficient of variation (%)
All (n=77)
Min–Max
Average7STD
Coefficient of variation (%)
a
Chlorophylls
Carotenoids
Flavonols
1.21–2.13
1.7770.29
16.4
0.39–4.01
2.0171.08
53.7
8.94–217
74.3770.5
94.9
0.71–2.03
1.2570.34
27.2
0.60–3.55
1.8270.64
35.2
13.9–176
57.9744.3
76.5
0.35–7.94
4.7472.81
59.3
0.78–3.69
2.3870.98
41.2
19.7–219
120769.9
58.3
0.35–7.94
2.5671.99
77.7
0.39–3.69
1.6870.73
43.5
8.93–219
85.6767.9
79.3
STD: standard deviation.
contained very low amounts of anthocyanins
(o0.15 nmol/cm2).
Fig. 2 shows reflectance spectra of apple fruits
differing in Flv content. All apple fruits possessed
high reflectance (up to 85%) in the NIR region of the
spectrum and lower reflectance in the red with a
distinct minimum near 678 nm, which is characteristic of Chl a (Fig. 2A). At wavelengths shorter than
550 nm, reflectance underwent a sharp decrease
due to combined absorption by Chl and Car. A
minimum near 480 nm attributable mainly to Car
and, to a lesser extent, Chl b absorption was
characteristic of the spectra of Golden Delicious
and Antonovka as well of sunlit surfaces of Renet
Simirenko fruits (Fig. 2). Shaded sides of Renet
Simirenko apples with high peel Chl (Fig. 2C,
curve 1) exhibited low and featureless reflectance
in the blue range (for more details of Chl and Car
spectral features in the visible range see Merzlyak
et al., 2003a). As a result of fruit adaptation to
strong sun-light, reflectance in the red range
increased, whereas that in the range 400–420 nm
decreased (Fig. 2). In fruits with high Flv content
reflectance in the range 400–420 nm was notably
lower than in fruits with low Flv content (cf. curves
1 and 2 in Fig. 2).
The effect of Flv on reflectance spectra in UV
range (300–400 nm) was studied in Antonovka fruits
grown locally and available at different stages of
their development (Fig. 3). The accumulation of Flv
in apple peel induced a gradual decrease of
reflectance in the range of 350–430 nm; these
spectra as well as corresponding spectra of reciprocal reflectance, 100/Rl, and their derivatives
are plotted in Fig. 3. At low Flv content the
reciprocal reflectance showed an increase from
400 nm toward shorter wavelengths (Fig. 3, curves
1 and 2). The rise in Flv content brought about the
appearance of a broad band between 360 and
420 nm with a remarkable shift of its edge towards
longer wavelengths (Fig. 3B, curves 3–5). In the
peel with the range of Flv content up to 120 nmol/
cm2, the relationship between inflection point
position, IP (maximum of the first derivative), and
Flv was linear (inset in Fig. 3C, Table 2). The
accumulation of Flv in higher amounts did not
result in further changes in IP position but caused
an increase in the peak height (cf. spectra 1–3 and
4, 5 in Fig. 3C). At wavelengths longer than 435 nm
the effect of an increase in Flv content was not
significant (Figs. 3B, C).
The spectra of correlation coefficient, r, of the
linear relationship between pigment content and
reciprocal reflectance for fruits of three apple
cultivars are shown in Fig. 4. Reciprocal reflectance
in the NIR did not relate to content of any pigment.
In the 530–730 nm range, the reciprocal reflectance
possessed a negative correlation with Flv content
(Fig. 4A), which was strong in apples with high-Chl
content (Renet Simirenko). In this spectral region,
100/Rl showed a high positive correlation with Chl
and Car. At shorter wavelengths (440–525 nm)
maxima attributable to Car absorption occurred
being especially pronounced in Antonovka and
Golden Delicious fruits. Prominent features of r
spectra were detected in the range 440–360.
Correlations of both Chl and Car content with
reciprocal reflectance underwent a sharp decrease
ARTICLE IN PRESS
Apple flavonols during fruit adaptation to solar radiation
155
Figure 3. Reflectance spectra (A) of Antonovka apple
fruits with different peel flavonol contents (1–45.7,
2–108.5, 3–121.8, 4–143.5; 5–233.8 nmol/cm2); (B) and
(C) corresponding reciprocal reflectance spectra and
their first derivatives, respectively.
Figure 2. Representative reflectance spectra of apples
with low (1) and high (2) flavonol content taken from
shaded and sunlit fruit surfaces, respectively. Pigment
contents are given in nmol/cm2.
at wavelengths shorter than ca. 450 nm (Figs. 4B,
C). In contrast, the correlation between Flv content
and reciprocal reflectance increased towards the
violet region with a peak around 410–420 nm. In the
blue (between 430 and 500 nm), the region of
combined absorption of Chl and Car, reflectance
showed low correlation with Flv regardless of their
content (Fig. 4A). The r spectra for 100/Rl vs. [Flv]
exhibited a distinct minimum at 500 nm and, in
Antonovka and Golden Delicious apples, also minima at 440 and 470 nm resulting from interference
exerted by Chl and Car (Figs. 4B, C). At wavelengths below 450 nm, the correlation of reciprocal
reflectance with Flv content underwent a sharp
increase and formed a conspicuous maximum
ARTICLE IN PRESS
156
M.N. Merzlyak et al.
Table 2. Algorithms for non-destructive determination of flavonol content in apple fruits with Flavonol Reflectance
1
Index, FRI: x ¼ ðR1
410 R460 ÞR800
Cultivar
Flavonol range (nmol/cm2)
n
Model
r2
RMSEa (nmol/cm2)
Golden Delicious
Antonovka
Renet Simirenko
All
8.9–217
13.9–176
19.7–219
8.9–219
13
35
29
77
½Flv ¼ 28:99 þ 14:63x
½Flv ¼ 1:66 þ 14:22x
½Flv ¼ 19:96 þ 13:64x
½Flv ¼ 5:08 þ 12:82x
0.96
0.92
0.92
0.92
16.2
12.9
20.8
19.3
a
RMSE: root mean square error of estimation.
corresponding troughs in 100/R(l)
(Figs. 4A, B) should be also noted.
vs.
[Chl]
Flavonol reflectance index
Based on the spectral characteristics of pigments,
we expected to find the optimal spectral regions l1,
1
l2, and l3 in the model ðR1
l1 Rl2 ÞRl3 for Flv
content estimation. The variability between samples in scattering can be eliminated by using
reflectance in a spectral region l3, where absorption by all pigments (Chl, Car, and Flv) is minimal
(i.e. in the NIR, l3 between 750 and 800 nm, see
Figs. 2–4). The only spectral range, l1, maximally
sensitive to Flv content was found in the blue
around 410 nm (Fig. 4). However, the correlation
between R800/R410 and Flv content was not significant (Fig. 5A) due to the effect of Chl and Car
absorption on reflectance (Figs. 2–4). To remove
the contribution of these pigments, reflectance at
l2 should be closely related to Chl and Car contents
and minimally affected by Flv absorption. In
addition, l2 should be as close as possible to l1.
The spectral range that meets the requirements
was assumed to be around 460 nm in the region
influenced mostly by combined Car and Chl
absorption and less affected by that of Flv (Fig. 4,
see also Merzlyak et al., 2003b). The Flavonol
Reflectance Index (FRI) was suggested in the form:
Figure 4. Spectral dependencies of correlation coefficient, r, between reciprocal reflectance, 100=Rl ; and
flavonol (A), chlorophyll (B) and carotenoid (C) content
for Golden Delicious (1), Antonovka (2), and Renet
Simirenko (3) apple fruits.
(r ¼ 0:9) at 410 nm (Fig. 4A). At the same time,
100/Rl in this spectral range demonstrated a
negative correlation with Chl and Car (Figs. 4B,
C). The presence of peaks around 530 nm in the
relationships 100/R(l) vs. [Flv] (more pronounced
in fruit varieties with lower Chl content) and
1
FRI ¼ ðR1
410 R460 ÞR800 :
Fig. 5B shows that subtraction of (R460)1 from
(R410)1 significantly decreased the effects of the
residual absorption by Car and Chl on reflectance
around 410 nm (Fig. 5B). FRI allowed accurate
assessment (r2=0.92, maximal RMSE=5.0 nmol/
cm2) of peel Flv content ranging from 8 to
220 nmol/cm2 for all apple fruit varieties studied
(Fig. 5B, Table 2). An estimation of Flv in the whole
range of its content required a variety-specific
approach and turned feasible with the use of power
or exponential relationships between FRI and Flv
content (not shown).
ARTICLE IN PRESS
Apple flavonols during fruit adaptation to solar radiation
Figure 5. Relationships between reciprocal reflectance
at 410 nm (A), flavonol reflectance index FRI, ðR1
410 R1
460 ÞR800 ; (B), and peel flavonol content. In B solid line
represents the linear fit for flavonol content in the range
8–220 nmol/cm2.
Discussion
Similar to leaves, fruit optical properties are
determined by pigments distributed in different
cell structures and their overall content, local
concentration, interactions, and organization, as
well as by scattering. Anthocyanin-free apple fruit
contain chloroplast pigments (Chl and Car) as
principal pigments absorbing in the visible range.
In addition, in UV both cuticular and vacuolar
phenolics are able to exert a strong contribution to
absorption.
As a result of the development under strong sunlight, remarkable changes in the fruit pigments
157
content were noted: the decrease in Chl, the
accumulation of additional amounts of Car, and the
build up of gross amounts of Flv (Fig. 1, Merzlyak et
al., 2002). On the whole, peel Flv content was 1–2
orders of magnitude higher than that of Chl and Car
(Table 1, see also Merzlyak et al., 2002). The
quantitative consideration—the ca. 5-fold difference between sunlit and shaded surfaces (Table 1,
Awad et al., 2000; Merzlyak et al., 2002)—presumes
that the latter response, which requires high
metabolic activity, entails energetic costs, and is
evidently involved in the fruit adaptation to solar
irradiation, is an important mechanism for protection against excessive fluxes of solar radiation. As
can be seen in Fig. 1, the cases with lower Flv
content more often occurred in shaded peel,
whereas high Flv content were predominantly
associated with sunlit peel. It should be also noted
that higher levels of Flv have been recorded in
Golden Delicious and Renet Simirenko apples grown
in South Russia.
In fruits with low Flv content, reflectance
between 440 nm (maximum of Chl a absorption)
and 400 nm was flat (Figs. 2 and 3) and then showed
a monotonous decline reaching its minimum near
360 nm (Fig. 3) close to rutin absorption maximum
in solution (Markham, 1989; Solovchenko et al.,
2001). These spectral features in UV could, to a
certain degree, be related to optical properties of
cuticles: those isolated from shaded surfaces of
Antonovka apples contained a considerable part of
peel Flv as well as that of other phenolics and
contributed appreciably to whole-fruit reflectance
in this spectral range (Solovchenko and Merzlyak,
2003).
The accumulation of Flv occurring mainly in the
vacuoles of subcuticular cell layers of the peel
(Awad et al., 2000) was accompanied by a sharp
decrease in fruit reflectance and flattening of the
spectrum in the broad band between 360 and
420 nm (Fig. 3A). In this spectral range, even at
high Flv content whole-fruit reflectance did not
drop below 4–5%. This effect is attributable to
reflectance from surface fruit structures, particularly the cuticle, which contains relatively low
amounts of peel Flv and exhibits considerable
scattering. Our previous experiments showed that
reflectance of isolated apple fruit cuticles recorded
in UV-A range on the ‘black background’ was
comparable with that of whole fruit containing
high amounts of peel Flv (Solovchenko and Merzlyak, 2003). Accordingly, the magnitude of UV
reflectance of apple fruit with high Flv content
should be strongly dependent on the physical
properties of the cuticular layer, its thickness,
composition, and organization of its lipid polymers.
ARTICLE IN PRESS
158
In apple fruit, reflectance near 360 nm was
quickly saturated reaching its minimal values at
relatively low peel Flv content (Fig. 3A). The build
up of Flv brought about a broadening of the
absorption band and manifested itself as the
extension of the region with very low reflectance
from near UV to the violet (Figs. 2 and 3). An
increase in peel Flv content up to ca. 140 nmol/cm2
was accompanied by moving the edge of fruit
reflectance spectrum towards longer wavelengths
(Fig. 3B) with the shift in IP position from ca. 364 to
425 nm (Fig. 3C). Qualitatively similar changes in
reflectance have been observed in the leaves and
fruits in the presence of high amounts of Chl
(Gitelson et al., 1996) and anthocyanins (Merzlyak
et al., 2002) in the red and green regions of the
spectrum, respectively. The Flv-dependent shift of
the edge of reflectance spectrum in apple fruit was
as high as 60 nm (Fig. 3). In addition to the
concentration-dependent effect of Flv on spectral
reflectance, the changes observed could be related
with their intermolecular interactions such as
copigmentation and aggregation of vacuolar Flv
resulting in a considerable bathochromic shift of
their absorption band as suggested for explanation
of yellow coloration exhibited by flower petals of
certain plant species (Smith and Markham, 1998;
Markham et al., 2001). This mechanism is quite
possible since local concentration of Flv in vacuoles
of apple peel cells is extremely high, reaching
1.7 102 M (Lancaster et al., 1994). As a result of
accumulation of high amounts of Flv the spectral
features of Chl a in the Soret band were masked by
Flv absorption both in whole apple fruit reflectance
(Figs. 2 and 3A, curves 4 and 5) and in absorption
spectra of peel extracts (Solovchenko et al., 2001).
The influence of high Flv content on fruit reflectance in the band 350–430 nm is also evident in the
spectra of correlation coefficients: the increase of r
for Flv coincided with a sharp decrease of r both for
Chl and Car (Fig. 4). Then, distinct positive peaks
of correlation between 100=Rl and Flv content
(Fig. 4A) were observed even in the region of high
reflectance in the green (maxima near 530 nm) as
well as corresponding negative peaks in the case of
Chl (Fig. 4B) and Car (Fig. 4C). The presence of
these features suggests that Flv in anthocyanin-free
fruits exert a contribution to light absorption and
hence to the screening of solar radiation not only in
UV-A but also in the visible part of the spectrum.
The non-destructive assessment of Flv in plant
tissues, which exhibit low reflectance in the UV-A is
complicated by overlapping of their absorption
with those of several pigments: Chl and Car, as
well as by phenolics (catechins and phenolic acids)
possessing absorption bands in UV-B (Krauss et al.,
M.N. Merzlyak et al.
1997; Awad et al., 2000). In addition, scattering
should exert a strong influence on UV reflectance.
Recently it was reported that scattering coefficients of whole fruit (Cubeddu et al., 2001), the
peel, and isolated cuticles of apples (Solovchenko
and Merzlyak, 2003) are wavelength-dependent
and underwent an increase with a corresponding
wavelength decrease.
In our previous studies, conceptual models have
been devised allowing non-destructive assessment
of total Chl content in higher plant leaves and fruits
(Gitelson et al., 2003; Merzlyak et al., 2003a).
Essentially the same models have been successfully
used for the non-destructive analysis of Car and
anthocyanin content (Gitelson et al., 2001, 2002,
2003; Merzlyak et al., 2003a, b). To apply the
conceptual model for quantitative estimation of Flv
content required finding optimal wavelengths l1, l2
and l3. At wavelengths shorter than 380 nm, the
reciprocal reflectance of apple fruits showed a
weak correlation with Flv content (Fig. 4A). This
could result from the saturation of the relationship
100/Rl vs. [Flv] as well as an interference by
phenolic substances different from quercetin glycosides (e.g. catechins and phenolic acids) (Awad et
al., 2000; Krauss et al., 1997). The band of the
highest sensitivity of reflectance to Flv in the whole
range of its changes has been found between 380
and 420 nm, peaking near 410 nm (Fig. 4). The
reflectance ratio R800/R410 possessed a low sensitivity to Flv content (Fig. 5A) due to a strong effect
of Chl and Car absorption. To remove their effect,
we suggested using reciprocal reflectance in a
spectral band that is closely related to the content
of the pigments. The spectral analysis revealed the
suitable band around 460 nm, where reflectance is
mainly governed by both Chl and Car (Figs. 2–4)
and, in addition, the correlation of reciprocal
reflectance with Flv content (Fig. 4) is low (Golden
Delicious and Antonovka) or negative (Renet Simir1
enko). The FRI in form ðR1
410 R460 ÞR800 exhibited a
strong variety-independent linear relationship with
Flv content in the range of 8–220 nmol/cm2 (Table
2). However, the sensitivity of FRI to Flv decreased
when Flv content exceeded 300 nmol/cm2 (Fig. 5B).
Thus, we spectrally tuned a conceptual model
developed for terrestrial plant leaves, and assessed
accurately Flv content in apple peels. The wide
range of optical properties of the fruits supports
the robustness of the findings. Our results provide
evidence that (a) the conceptual model can be
applied for accurate non-destructive estimation of
Flv content in apples, and (b) fine tuning of the
conceptual model can be carried out knowing the
spectral characteristics of the specific medium of
interest.
ARTICLE IN PRESS
Apple flavonols during fruit adaptation to solar radiation
It should be noted in conclusion that the
influence of Flv on optical spectra of apple fruits
(Figs. 2–4) and higher plant leaves might extend
quite far into the visible spectrum. Therefore, one
using reflectances for non-destructive determination of higher plant pigments absorbing in the
visible range should be aware of obstacles that
could be caused by Flv when they are present in
high amounts.
References
Awad MA, de Jager A, van Westing LM. Flavonoid and
chlorogenic acid levels in apple fruit: characterisation
of variation. Sci Hort 2000;83:249–63.
Bornman JF, Reuber S, Cen Y- P, Weissenböck G.
Ultraviolet radiation as a stress factor and the role
of protective pigments. In: Lundse J editor. Plants
and UV-B: Responses to environmental change.
Cambridge, New York: Cambridge University Press;
1997. p. 157–68.
Burchard P, Bilger W, Weissenböck G. Contribution of
hydroxycinnamates and flavonoids to epidermal
shielding of UV-A and UV-B radiation in developing
rye primary leaves as assessed by ultraviolet-induced
chlorophyll fluorescence measurements. Plant Cell
Environ 2000;23:1373–8.
Cerovic ZG, Ounis A, Cartelat A, Latouche G, Goulas Y,
Meyer S, Moya I. The use of chlorophyll fluorescence
excitation spectra for the non-destructive in situ
assessment of UV-absorbing compounds in leaves.
Plant Cell Environ 2000;25:1663–76.
Cubeddu R, D’Andrea C, Pifferi C, Taroni P, Torricelli A,
Valentini G, Ruiz-Altisent M, Valero C, Ortiz C, Dover
C, Johnson D. Time-resolved reflectance spectroscopy
applied to the non-destructive monitoring of the
internal optical properties in apples. Appl Spectrosc
2001;55:1368–74.
Day TA, Martin G, Vogelmann TC. Penetration of UV-B
radiation in foliage: evidence that the epidermis
behaves as a non-uniform filter. Plant Cell Environ
1993;16:735–41.
Escarpa A, González MC. High-performance liquid chromatography with diode-array detection for the determination of phenolic compounds in apple skin and pulp
from different apple varieties. J Chromatogr A 1998;
823:331–7.
Gitelson AA, Gritz U, Merzlyak MN. Relationships between
leaf chlorophyll content and spectral reflectance and
algorithms for non-destructive chlorophyll assessment
in higher plant leaves. J Plant Physiol 2003;160:
271–82.
Gitelson AA, Merzlyak MN, Chivkunova OB. Optical
properties and non-destructive estimation of anthocyanin content in plant leaves. Photochem Photobiol
2001;74:38–45.
Gitelson AA, Merzlyak MN. Signature analysis of
leaf reflectance spectra: algorithm development for
159
remote sensing of chlorophyll. J Plant Physiol
1996;148:495–500.
Gitelson AA, Merzlyak MN, Lichtenthaler HK. Detection
of red edge position and chlorophyll content by
reflectance measurements near 700 nm. J Plant
Physiol 1996;148:501–8.
Gitelson AA, Zur Y, Chivkunova OB, Merzlyak MN.
Assessing carotenoid content in plant leaves with
reflectance spectroscopy. Photochem Photobiol 2002;
75:272–81.
Harborne JB, Williams CA. Advances in flavonoid research
since 1992. Phytochem 2000;55:481–504.
Havaux M, Kloppstech K. The protective functions of
carotenoids and flavonoid pigments against excess
visible radiation at chilling temperature investigated
in Arabidopsis npq and tt mutants. Planta
2001;QJ;213:953–66.
Jansen MAK, Gaba V, Greenberg BM. Higher plants and
UV-B radiation: balancing damage, repair and acclimation. Trends Plant Sci 1998;3:131–5.
Kolb CA, Käser MA, Copeckỳ J, Zotz G, Riederer M,
Pfündel EE. Effects of natural intensities of visible and
ultraviolet radiation on epidermal ultraviolet screening and photosynthesis in grape leaves. Plant Physiol
2001;127:863–75.
Krauss P, Markstädter C, Riederer M. Attenuation of UV
radiation by plant cuticles from woody species. Plant
Cell Environ 1997;20:1079–85.
Lancaster JE, Grant JE, Lister CE, Taylor MC. Skin color in
apple—influence of copigmentation and plastid pigments on shade and darkness of red color in five
genotypes. J Am Soc Hortic Sci 1994;119:63–9.
Liakoura V, Bornman JF, Karabourniotis G. The ability of
abaxial and adaxial epidermis of sun and shade leaves
to attenuate UV-A and UV-B radiation in relation to the
UV absorbing capacity of the whole leaf methanolic
extracts. Physiol Plant 2003;117:33–43.
Markham KR. Flavones, flavonols and their glycosides.
Harborne JB, Dey PM editors. Methods in Plant
Biochemistry, Vol. 1. London: Academic Press; 1989.
p. 197–232.
Markham KR, Gould KS, Ryan KG. Cytoplasmic accumulation of flavonoids in flower petals and its relevance to
yellow flower colouration. Phytochemistry 2001;58:
403–13.
Mazza CA, Boccalandro HE, Giordano CV, Battista D,
Scopel AL, Ballaré CL. Functional significance and
induction by solar radiation of ultraviolet-absorbing
sunscreens in field-grown soybean crops. Plant Physiol
2000;122:117–25.
Merzlyak MN, Gitelson AA, Chivkunova OB, Solovchenko
AE, Pogosyan SI. Application of reflectance spectroscopy for analysis of higher plant pigments. Russ
J Plant Physiol 2003a;50:704–10.
Merzlyak MN, Solovchenko AE, Chivkunova OB. Patterns
of pigment changes in apple fruits during adaptation
to high sunlight and sunscald development. Plant
Biochem Physiol 2002;40:679–84.
Merzlyak MN, Solovchenko AE, Gitelson AA. Reflectance
spectral features and non-destructive estimation of
ARTICLE IN PRESS
160
chlorophyll, carotenoid and anthocyanin content in
apple fruit. Postharv Biol Technol 2003b;27:88–103.
Reay PF, Lancaster JE. Accumulation of anthocyanins and
quercetin glycosides in ‘Gala’ and ‘Royal Gala’ apple
fruit skin with UV-B-Visible irradiation: modifying
effects of fruit maturity, fruit side, and temperature.
Sci Hortic 2001;90:57–68.
Smith GJ, Markham KR. Tautomerism of flavonol glucosides:
relevance to plant UV protection and flower colour. J
Photochem Photobiol A: Chem 1998;118:99–105.
Solovchenko A, Merzlyak M. Optical properties and
contribution of cuticle to UV protection in plants:
experiments with apple fruit. Photochem Photobiol Sci
2003;2:861–6.
Solovchenko A, Schmitz-Eiberger M. Significance of skin
flavonoids for UV-B protection in apple fruits. J Exp
Bot 2003;54(389):1977–84.
M.N. Merzlyak et al.
Solovchenko AE, Chivkunova OB, Merzlyak MN, Reshetnikova IV. Spectrophotometric pigment analysis in apple
fruit. Russ J Plant Physiol 2001;48:693–700.
Strack D, Wray V. Anthocyanins. Harborne JB, Dey PM
editors. Methods in Plant Biochemistry, Vol. 1. New
York: Academic Press; 1989. p. 326–52.
Takahama U. Redox reactions between kaempferol
and illuminated chloroplasts. Plant Physiol 1983;71:
598–601.
Tevini M, Braun J, Fieser G. The protective function of
the epidermal layer of rye seedlings against ultraviolet-B radiation. Photochem Photobiol 1991; 53:
329–333.
Wellburn AR. The spectral determination of chlorophylls
a and b, as well as total carotenoids, using various
solvents with spectrophotometers of different resolution. J Plant Physiol 1994;144:307–13.